Sink or SWIN? Image Similarity Search with ViT to Stop Phishing Attacks

Logo
Presented by

Adithya Singh, Machine Learning Engineer at Bolster

About this talk

Vision Transformers are quickly emerging as the top method in solving computer vision tasks for anti-phishing use cases due to the immense success of transformers in the NLP domain. For the longest time, CNNs and hashing systems have been go-to methods to solve image classification problems. However, many limitations of these methods have given rise to powerful vision transformers like SWIN which is taking the field of computer vision by storm. Join us as we do a technical deep dive into how image similarity search to solve image classification and remarkable capabilities of SWIN transformers, which leverage the power of self-supervised learning and visual recognition to identify sophisticated phishing campaigns. Adithya Singh, Machine Learning Engineer at Bolster, will discuss how SWIN transformers can be used for: - Analyzing complex visual cues for threat categorization - Detecting subtle manipulations in the visual data - Distinguishing with high fidelity legitimate websites from malicious ones; - and being added to your existing tools for phishing and scam detection
Related topics:

More from this channel

Upcoming talks (1)
On-demand talks (11)
Subscribers (1154)
Thought provoking webinars featuring cybersecurity subject matter experts discussing recent trends on digital risk protection services, phishing protection, threat intelligence, artificial intelligence, machine learning, social media monitoring, external security, trademark infringement, scam and fraud protection, security operations, dark web monitoring, brand protection, and more! The channels is educational for for CIOs, CISOs, S/VPs, Directors, Managers, Architects, Engineers, SecOps and Analysts in Security and IT.